Optical Sensor Evaluation for Vision Based Recognition of Electronics Waste

Abstract

For the online and offline classification of printed circuit boards and SMD components mounted on the board a sensor evaluation of optical sensors with different characteristics is presented. In overall a high resolution multispectral camera, a DSLR, an IP camera, and the RGB sensor of the Asus Xtion Pro Live have been evaluated for 2D images. For the 3D acquisition the Asus Xtion Pro Live is compared to the Bluetechnix M100 time of flight sensor. The sensor analysis shows the affected recognition performance of printed circuit boards and SMD components. Based on the sensor comparison, a pilot plant for the recycling of electronics waste is presented.